Economics - Home Scholars At Harvard

2y ago
2 Views
2 Downloads
230.14 KB
40 Pages
Last View : 23d ago
Last Download : 2m ago
Upload by : Mya Leung
Transcription

Tirole’s Industrial Regulation and Organization Legacy inEconomicsDrew FudenbergMarch 21, 2015I thank Glenn Ellison, Mira Frick, Ryota Iijima, Eric Maskin, Lones Smith, Patrick Rey, JeanCharles Rochet, Al Roth, Nathalie Tirole, and Jean Tirole for helpful comments. NSF grantSES-1258665 provided financial support.

1.Introduction“Je suis un chercheur; je ne suis pas capable de parler de toutet n’importe quoi.”1Jean Tirole was awarded the 2014 Sveriges Riksbank Prize in Economic Sciences inMemory of Alfred Nobel for his analysis of market power and regulation. This essay will try toconvey the main ideas of that work, along with some of Tirole’s positive and methodologicalcontributions to the study of imperfect competition and its implications for industrialorganization. This narrow focus emphasizes my connections with Jean, and leaves out his manyimportant contributions in such fields as asset pricing, behavioral economics, and organizationaleconomics, which on their own would constitute a stellar career, as well as his arguably Nobellevel work on banking and corporate finance.2 Jean’s phenomenal energy and breadth arereflected in the fact that he has distinct and influential collaborations with each of PhilippeAghion, Roland Benabou, Mathias Dewatripont, Oliver Hart, Bengt Holmstrom, Paul Joskow,Jean-Jacques Laffont, Josh Lerner, Eric Maskin, Patrick Rey, and Jean-Charles Rochet. Any ofthem could have been asked to write this essay, and each would have their own take on Jean’sstory; I would like to share a little of my own.I first met Jean in 1978 when we both started graduate school at MIT. At the time,theorists such as Dasgupta, Dixit, Spence, and Stiglitz were exploring the implications ofdynamic issues such as commitment and timing for such industrial organization topics as patent1Tirole explaining why he won’t offer opinions on arbitrary topics: “ I am a researcher, and not able to talk abouteverything and anything” : t-journal/pid7560vu.html?vid 1148121. While quite a reasonable position, Jean’s areas of competence are much larger than those ofmost economists I know.2For a broader and more detailed overview of Jean’s work, see the Economic Sciences Prize Committee’sexcellent scientific report; I relied on it extensively when writing this essay. Another sign of what this essay leavesout is the fact that the 2015 Nemmers Prize conference in his honor, “Liquidity, Bubbles, and Crises,” focused onJean’s work on banking and financial markets as opposed to my focus on regulation and IO.

races and preemptive investment.3 Simultaneously, game theorists were developing tools thatseemed natural for studying these problems, such as sequential equilibrium (Kreps and Wilson(1982)), and showing how to use formal game theoretic models to study dynamic competition, asin Milgrom and Roberts (1982) on limit pricing. Neither game theory nor mechanism designwas then a standard part of the economics curriculum, but Eric Maskin taught an advanced classon these topics.4 As important, Eric spent a lot of additional time with us in a reading class wherewe read a number of soon-to-be classic papers, including all of the game theory papersmentioned above. Jean’s attraction to game theory, and particularly dynamic games, was quickto take root: During graduate school he analyzed the strategic aspects of capacity expansion,learning-by-doing, and bargaining with incomplete information (Fudenberg and Tirole (1983a),(1983b), (1983c)), and began working on the dynamic oligopoly models that became Maskin andTirole (1987), (1988a,b). 5Since then, Jean has been a leader in applying game theory and mechanism design toanalyze how firms set prices, make investment decisions, etc., and how to design rules andregulations that lead to better outcomes. As the Nobel committee noted, “No other scholar hasdone more to enhance our understanding of IO in general, and of optimal policy interventions inparticular.” He did this by identifying important economic problems, developing and extendingthe appropriate game theoretic tools, and applying them to derive important conclusions andresults. After Jean moved to Toulouse in 1991, he and Laffont began working with Electricitéde France and France Telecom, and those connections would start a fruitful feedback loop3See for example Spence (1977) and Dixit (1980) on strategic capacity expansion, and Dasgupta and Stiglitz(1980) on patent races.4The syllabus for the class was the minmax theorem, Nash equilibrium (including existence in discontinuous games,and applications to implementation theory), the core, Shapley value, bargaining set, and other cooperative solutionconcepts.5In addition to all of this, Jean found time to write a paper on fixed-price equilibria (Maskin and Tirole (1984)).

between research and its application, but it is worth bearing in mind that in Jean’s case theresearch interest came first.The move to Toulouse brings up another major part of Jean’s contribution to economics,namely helping to build an excellent economics group in a French university. The first steptowards this program was the establishment of the IDEI (Institut d’Economie Industrielle),which Jean helped Laffont found in 1990. Jean has played a lead role in the growth of the IDEI,and in the creation and success of the Toulouse School of Economics, which has become one ofthe best economics groups in Europe. This institution-building has many positive externalitiesfor the profession, and Jean is rightly very proud of it. The time and effort involved make hisresearch accomplishments all the more impressive.Finally, while the Nobel Prize is not awarded for textbooks, Jean’s contribution in thisdomain are far-reaching and important. To quote the Prize Committee, “After more than 25years, his groundbreaking 1988 textbook remains the best road-map to the field. If the book isbecoming outdated in a few areas, this is largely due to Tirole’s own subsequent work and thework he has inspired. Tirole’s 1993 book, co-authored with Jean-Jacques Laffont, presented aunified framework which has deeply influenced how economists think about regulation.” Theseare only two of Jean’s twelve books; I say more about a few of them below.2.Regulation as Applied Mechanism Design2A: The Regulation of MonopoliesMany governments use regulations to reduce the distortions that would be caused bymonopoly pricing, and there is a long tradition of work asking what sorts of regulations shouldbe used. Ramsey (1927), Boiteux (1956) and others studied how to set prices to maximize totalsocial surplus (that is, consumer surplus plus firm surplus plus taxpayer surplus, ignoring

distributional concerns) given a break-even constraint in static settings where the regulatorknows both the production technology and the demand function. Here the optimal policy is toset the proportional markup ( pi ci ) / pi equal to θ / ηi , where ηi is the elasticity of demand ofgood i and θ [0,1] reflects the tightness of the break-even constraint or the social cost ofproviding subsidies. Thus there is less price distortion on goods where markups cause moredeadweight loss, and the ratio between the regulated markup and the monopoly markup is thesame for every good. In practice, though, regulators may not know the firm’s production cost,and this sort of cost-plus pricing does not provide the correct incentives for cost-reducinginvestment. This may be why regulated firms have often been governed by “rate of return”regulation, which sets prices so that the firm earns a “reasonable” return on its investments.Some of the early work on the theory of regulation, such as Averch-Johnson (1962),analyzed how to set the parameters of commonly-used regulatory rules. This did provide someinsights on the expected distortions that these rules cause, but since these rules were taken asgiven, without an explanation of when or why they would be used, the analysis was incomplete:Might there be other simple and practical rules that addressed the underlying incentive problemsand led to a better outcome?Baron and Myerson (1982) pioneered the modern study of regulation- analyzing the fullyoptimal direct mechanism in a setting with all of the objectives and constraints made explicit- intheir analysis of regulation with adverse selection, where the monopolist’s cost is unknown butfixed and can’t be altered by the firm’s actions. 6 Sappington (1982) allowed for moral hazard in6Vogelsang and Finsinger (1979) and Loeb and Magat (1979) had previously incorporated asymmetric informationinto models of regulation, but each made key simplifications that limited the impact of their work: Vogelsang andFinsinger (1979) studied a dynamic process of rate adjustment when the firm responds completely myopically, andLoeb and Magat (1979) assumed that there is no social cost to the rents of the firm - here it is easy to implementthe efficient pricing rule by paying the firm the whole consumer surplus.

addition to adverse selection, as well, supposing that the firm can do unobserved research toreduce its production cost, but restricted attention to linear contracts.Using Cost Information to Regulate FirmsLike Sappington, Laffont and Tirole (1986) suppose that the monopoly firm is subject tomoral hazard as well as adverse selection, allowing for either perfect or noisy observation ofcosts. They then use the revelation principle to characterize the fully optimal menu of contracts.Myerson (1982) had shown that the revelation principle applies in these problems, but did notexplicitly analyze one; Laffont and Tirole (1986) appears to be the first to have provided anexplicit solution to one.In the simplest version of their model, a firm- identified with a single agent- produces asingle output q at cost c ( β e)q , where e 0 is the firm’s effort, and βis the firm’sefficiency parameter, which is known to the firm but not to the regulator.7 (The paper allows costto also depend on a mean-zero cost shock that is independent of β and does not depend on q ande, but this makes no difference for most of the analysis because both the firm and the regulatorare risk neutral and transfers can be arbitrarily large.) The good provides consumer surplusS (q ) , and to simplify, the paper assumes that the good is not marketed but provided at no costto consumers. 8 The planner observes and reimburses the cost c incurred by the firm and pays inaddition a net monetary transfer t so that the total payment is c t . The utility level of the firm’smanager is then U t ψ (e) , where ψ measures the disutility of effort.7The identification of the firm with a single agent is a standard assumption in most of the regulation and industrialorganization literatures, which gloss over the internal structure of the firm, but see Chapter 7 of Tirole (2006) for adiscussion of how that structure can interact with behavior in the product market. One way to interpret the firm’s“effort” here is as a (restraint on) wasteful perks for management.8In a footnote, the paper explains that modelling private sales just requires adding the social value of the associatedrevenue to the objective function. This case was studied in the working paper version, and is explored at length inLaffont and Tirole (1993).

The regulator’s objective is to maximize the sum S U of consumer and producersurplus, minus the social cost (1 λ )(c t ) of the funds transferred to the firm, where λ 0reflects the shadow cost of the distortionary taxation. Thus, the regulator would prefer to makesmall transfers, but must balance this with concerns for efficient effort and socially optimalprovision of the good. The regulator knows all of the parameters of the model, except for thefirm’s type β ; it can set the output level q and it will observe the firm’s cost c.Using the revelation principle, the regulator’s problem is to design a direct mechanismunder which each report βˆ is assigned an effort level e( βˆ ) , anticipated per-unit cost c ( βˆ ) , andoutput q ( βˆ ) that satisfy the incentive-compatibility constraints that each type of firm prefer toannounce truthfully and take the assigned effort level, and also satisfy the interim individualrationality constraints that all types are willing to participate. Under some technicalassumptions,9 Laffont and Tirole (1986) showed that it is sufficient to consider only the localincentive constraint, and from there they proceeded to show that the optimal mechanism can beimplemented by having the firm pick from a menu of contracts, where instead of specifying aquantity, effort level, and transfer, each contract specifies a quantity, a lump-sum payment, andthe share of the realized costs that the regulator will reimburse. Moreover, the menu is such thatthe most efficient firm (type β ) chooses a fixed-price contract without any cost sharing at all, sothat it chooses the cost-minimizing level of effort given its assigned production: As in Mirrlees(1971) and subsequent work, when only the local incentive constraints bind there is “nodistortion at the top.” The planner could assign such contracts to all types, and indeed wouldhave to do so if costs were not observed, as in Baron and Myerson, but this would increase the9All of the functions above, as well as the density of types, are continuously differentiable, S is concave, ψ isconvex, and the density satisfies the monotone hazard rate property. Additional assumptions are used to ensure thatit is neither optimal to shut down the firm nor send the marginal cost of production to 0, that the full informationproblem is “sufficiently convex,” and that the first-order condition for the planner’s maximization is sufficient.

size of the payments needed to ensure truthful reporting. Instead, the optimal contract providescost sharing and lower lump sum payments to higher types, so in general the effort level is lessthan the optimum, costs are higher, and production is lower than with perfect information. Inaddition, the optimal contract moves toward a fixed-price contract when demand increases,because the importance of per-unit cost reduction is higher when more is produced.In subsequent work, Laffont and Tirole extended this model to analyze a range of relatedregulatory issues, still within the static framework. Laffont and Tirole (1990a and 1993 Chapter3) study multi-product firms, and show that when cost satisfies a separability condition, therelative prices obey the Ramsey-Boiteux formulae, so that the regulator doesn’t distort the pricestructure to try to extract rents; rent extraction is addressed through the regulation of the rate ofreturn. This result provides a foundation for the widespread use of regulations that imposeaggregate caps but let the firm allocate its constrained exercise of market power according todemand elasticities. Laffont and Tirole (1993, Chapter 4) extend the model to the (indirect)regulation of quality when quality is not verifiable and so cannot be included in the terms of themechanism, and Laffont and Tirole (1990c and 1993 Chapter 6) study “cream skimming” by afirm that faces two different markets or types of consumers.In hindsight, given Jean’s later work on platform competition and two-sided markets, themost evocative static extension may be Laffont and Tirole (1993 Chapter 5 and 1994) on accesspricing, which involves a regulated multi-product firm some of whose products (e.g. access to anetwork of optical or electricity cables) are sold to other firms who then compete with theregulated firm for consumers. If the firm’s cost of providing access to others is the same as thecost of using the network itself, the only way the firm could discourage access is by saying thatthe cost of the network is high, and this would lead the regulator to reduce planned supply, which

reduces the firm’s incentive to exaggerate cost. On the other hand the firm would be very keen toexaggerate any incremental cost that is needed solely to provide access to others. Typically, theregulator will choose to only partially offset this incentive, so the optimal mechanism will tendto have higher access prices and less competition than with full information.The Dynamics of RegulationWhile Laffont and Tirole (1986) and much of Laffont and Tirole (1993) studied a staticproblem, Laffont and Tirole (1988, 1990b, 1993 Chapters 9 and 10) studied dynamic versionsof their basic (1986) model, with the simplification that output is simply 0 or 1 (as if demand isvery inelastic, or if the firm is the sole plausible supplier of an indivisible public good such as abridge or train line.) Baron and Besanko (1984) had shown that when the regulator can committo future rules, and the firm’s type is the same in every period, the optimal dynamic scheme issimply to commit to enforcing the optimal static scheme in each period, much as the optimalpricing policy for a monopoly seller of a durable good is to commit to a fixed price (Stokey(1981)).10 In practice such commitment may be difficult to achieve, which motivated Laffontand Tirole to consider other scenarios. Laffont and Tirole (1988) supposes that the regulator isonly able to offer short-term contracts, so that not only can the regulator not commit, the firmcannot either, as it has the option of rejecting the current contract. The regulator’s lack ofcommitment gives rise to a ratchet effect (as in Weitzman (1976)): when determining thecontract to offer in the last period of the relationship, the regulator will use all of the informationhe has acquired previously, but this makes the firm less willing to reveal information in earlierperiods.10They also considered the case where the type can change from period to period, a topic which has attractedincreased attention in recent years.

Freixas, Guesnerie, and Tirole (1985) considered this sort of problem in a pure adverseselection setting with a restriction to linear contracts; Laffont and Tirole (1988) considered theequilibria with an unrestricted contract space in a two-period version of the cost-observationmodel of Laffont and Tirole (1986). Here, the static optimum cannot be implemented, becausein addition to the usual problem in static models of deterring the more efficient types fromclaiming to be inefficient, the regulator needs to worry that a less efficient firm will claim to beefficient, collect a large lump sum (which is the optimal contract for an efficient agent in thestatic model) and then refuse the second period contract. That is, the incentive constraints canbind in both directions, which makes them difficult to characterize. Moreover, no period-1mechanism can induce full revelation of types- there will always be some degree of pooling, soafter period 1 the regulator will in general not know the firm’s true type.11 Laffont and Tiroleshow that when there is little uncertainty about the firm’s type (in the sense that the typedistribution is a small interval) then either almost all types pool together in the first period (sothat the lack of commitment leads to an almost complete breakdown of the informationrevelation obtained in the static case) or the equilibrium takes a very complex “non-partition”form which seems unlikely to occur in practice.In Laffont and Tirole (1990b), the regulator and firm can enter into a long-term contractwhich either of them can enforce- so the regulator can commit not to use first period informationto extract all of the firm’s second period rents, and the firm can commit itself not to “take themoney and run.” However, the two parties are unable to jointly commit not to tear up thecontract and sign a new one. That is, the contract must be renegotiation-proof in the sense of11If there were a separating equilibrium, a less efficient firm that claims to be more efficient than it is in the firstperiod will choose to exit in the second period, while a more efficient one that claims to be less efficient will earn apositive second-period surplus. The conclusion then follows from the constraints that neither firm wants to announcethe other’s type.

Dewatripont (1989). Here there are contracts that lead to the full separation of types, but this isnot optimal for the regulator, and the best contract involves some pooling. The paper establishesthis result when the prior is a smooth density over a continuum of types, but the detailed analysisrestricts to the two-type case. Here the second-period contract is conditionally optimal- meaningthat it is optimal in the continuation game that starts in the second period- though the rent giventhe good type depends on second period beliefs and hence on first-period play. In the firstperiod, the principal offers two contracts, one without any cost sharing (as in the contract for themost efficient type in Laffont-Tirole (1986)); the “bad” type accepts the other contract (whichdoes have some cost sharing) and the good type randomizes so that the equilibrium is “semiseparating”: accepting the fixed-price contract reveals the firm’s type, but accepting the contractwith cost sharing does not. Thus, as in the case of only short-term contracts, there is somepooling of types to reduce the regulator’s ex-post incentive to extract rents from the firm.12Regulatory CaptureStigler (1971) pointed out that regulatory agencies might be “captured” by those mostaffected by their regulations, and showed that trucking regulations in the 1930’s seemed to havebeen heavily influenced by the competing railroad industry. Peltzman (1976) took a steptowards formalizing the model, but left the behavior of consumers and firms exogenous, and didnot incorporate the informational asymmetries and other constraints that make the design ofregulations non-trivial and allow regulated firms to obtain the rents and thus give them anincentive to distort the process. To model regulatory capture, Laffont and Tirole (1991, 1993chapter 11) combined their basic “cost observation” model with the three-tier structure that12For the same reason, the optimal renegotiation-proof contracts are semi-separating in the Hart-Tirole (1988)model of renegotiation-proof contracts for the repeated rental of a durable good, and in the Fudenberg-Tirole (1990)model of renegotiation-proof contracts when the agent faces moral hazard and contracts can be renegotiated after theagent has made the effort decision but before all of its effects are realized.

Tirole (1986a) introduced to model collusion in general hierarchies. Here, as in the basic model,the firm has cost function C ( β e)q , where β is private information, and has only twopossible values, β β . Congress takes the role of the regulator in the base model, and has thesame payoff function, and observes the firm’s realized cost. In addition, there is a supervisorwho with some probability learns the firm’s cost parameter β . The supervisor then sends a reportof its information to Congress. Importantly, when the supervisor does receive a signal it isverifiable, as in Grossman (1981) and Milgrom (1981) - the supervisor cannot falsely report thatthe firm’s type is high ( β ) when it is low ( β ). However, unlike in those papers, there is achance that the supervisor is uniformed (receives a null signal) and the supervisor can falselyclaim to be uninformed.13 In the absence of collusion with interested parties, none of thismatters, and the regulator reports truthfully, but if the supervisor learns that the firm’s cost islow, the firm has an incentive to bribe her to suppress that information. Foreseeing this, Congressknows it must pay the supervisor a bonus when she reports that the type is low, and since thesize of the bonus depends on the low type’s gain from reporting high, the optimal mechanismwhen these bribes are possible will sacrifice efficiency to reduce this gain and thus reduce thetransfers needed to generate honest reporting by the supervisor. In particular, the optimalregulatory scheme now further distorts the effort and output of the inefficient firm by using acontract with lower powered incentives (that is, more cost sharing, so that the transfer is less13In practice supervisors and regulators cannot simply choose to not send a report, so reporting the null messageshould be thought of as leaving some relevant information out of the report. In Grossman’s signaling model, anunravelling argument shows that failure to report will be interpreted as revealing the worst possible information; thatconclusion extends to sufficiently small probabilities that the sender is uninformed.

responsive to the outcome) as in Tirole (1986a), which can be seen as a defense or explanationfor the perhaps surprising prevalence of such contracts.142BRegulation of OligopoliesVertical restraintsVertical relationships between firms, where one firm sells to another, often usecontractual restrictions such as resale price maintenance or exclusive territories. Theserestrictions have sparked a large number of anti-trust cases, and a sizable economic literature.As with the analysis of price regulation, Tirole’s approach was to explicitly model thecontracting problems that prevent the participants from obtaining their preferred outcome. In Reyand Tirole (1986), a monopoly upstream firm sells to a number of geographically dispersedretailers. The contracting problem arises because the (location specific) cost and demand areuncertain, and retailers observe these local shocks but the monopolist does not. In addition, themonopolist is unable to observe actual sales of each retailer, so to avoid arbitrage it must set thesame variable price to all of them.15 In this case the monopolist can effectively only use two-partpricing.In the absence of vertical restraints the resulting Bertrand competition leads the retailprice to be independent of local shocks to demand and to respond fully to local cost shocks, sothat the equilibrium franchise fee (the fixed part of the monopoly price) is zero.Rey and Tiroleshow that the monopolist does better by assigning each retailer an exclusive region in exchangefor a positive franchise fee provided that the retailers are not too risk averse. This allows the14The extent of this distortion depends on the probability that the supervisor is uninformed, and vanishes as thisprobability goes to 0. Laffont and Tirole also consider an extension of the model with “interest groups” who canspend resources to influence the supervisor’s report, and show that the ability to do so can make the groups worseoff when it is foreseen by the mechanism designer.15The monopolist also wants (or is required) to sell to all of the retailers, so it can’t auction off rights to all of themarkets to one retailer.

retail price to respond to demand shocks and to respond less to cost shocks, so that moreconsumer surplus can be extracted. When retailers are risk-neutral, the monopolist gains fromexclusive dealing, even though it lowers social welfare, and the same is true with linear demandwhen retailers aren’t too risk averse. This shows that the private optimality of exclusive dealingis not prima facie evidence that it is socially desirable.Instead of asymmetric information, Hart and Tirole (1990) consider the problems posedby the upstream monopolist’s inability to make binding commitments. They assume a knowndemand curve, so that with full commitment the monopolist can obtain the monopoly profit witha “forcing” contract that specifies the quantity that each retailer will sell, charging them a lumpsum equal to their earnings. Hart and Tirole’s insight is that this contract is not an equilibrium ifthe monopolist can offer secret price cuts or otherwise renegotiate with any individual retailer,as the two of them can extract profit from the others by slightly increasing that retailer’s output.16However, if the monopolist can credibly commit to an exclusive dealing arrangement with asingle retailer (which was ruled out by assumption in Rey and Tirole), or can vertically integratewith one of the retailers, then the problem with secret side-deals goes away and the upstreamfirm receives the monopoly profit, as it does if the government enforces a rule that themonopolist cannot discriminate between retailers.TelecommunicationsIn the 1960’s and 70’s, telecommunications in the US and many other countries was runby either a private but regulated monopoly or a state-run firm, but since the early 80’s the trendhas been towards privatization and oligopoly. Because telecoms need to cooperate to send16The easiest way to see this is may be to suppose that the monopolist doesn’t actually produce anything, but is amediator who is empowered to set output limits for each retailer. The retailers can collude on the monopoly output ifthe mediator can make binding commitments, but absent commitment power the collusive agreement would unravelto Cournot competition.

messages from one network to another, this raises a number of new regulatory issues related toaccess charges (e.g. when one firm has a monopoly on a trunk or intercontinental line) andinterconnection or termination fees (when a consumer calls a customer of another network).Laffont, Rey and Tirole (1998a, 1998b) developed a formal model of two-way access, based onthe assumptions that the receiver does not pay for calls (the caller’s company pays terminationfees); Laffont et al (2001) extend this to allow charges to both sides. Laffont and Tirole (2000)consider the optimal regulation of termination and access fees, and argue that the consumer sideof the market is best regulated with price caps.Joint Marketing and Patent PoolsSuppose that n firms each have 1 patent, which they could sell separately. When andunder what conditions should they be allowed to form a patent pool that sells all of the patents asa single bundle? Shapiro (2001) pointed out that the static Cournot model implies that pools areanti-competitive (lead to higher prices and lower sales) when the p

learning-by-doing, and bargaining with incomplete information (Fudenberg and Tirole (1983a), (1983b), (1983c)), and began working on the dynamic oligopoly models that became Maskin and Tirole (1987), (1988a,b). 5 Since then, Jean has been a leader in appl

Related Documents:

Life science graduate education at Harvard is comprised of 14 Ph.D. programs of study across four Harvard faculties—Harvard Faculty of Arts and Sciences, Harvard T. H. Chan School of Public Health, Harvard Medical School, and Harvard School of Dental Medicine. These 14 programs make up the Harvard Integrated Life Sciences (HILS).

Harvard Business School Department of Economics, Harvard University 617-495-6644, asunderam@hbs.edu 617-496-2614, maggiori@fas.harvard.edu Teaching and Research Fields: Primary fields: Financial Economics, Corporate Finance . -1- _@HARVARD.EDU Placement Director: .

Std. 12th Economics Smart Notes, Commerce and Arts (MH Board) Author: Target Publications Subject: Economics Keywords: economics notes class 12, 12th commerce, 12th economics book , 12th commerce books, class 12 economics book, maharashtra state board books for 12th, smart notes, 12th std economics book , 12th economics book maharashtra board, 12th economics guide , maharashtra hsc board .

International Finance 14. Development Policy 15. Institutional Economics 16. Financial Markets 17. Managerial Economics. 13 18. Political Economy 19. Industrial Economics 20. Transport Economics 21. Health Economics 22. Experimental and Behavioral Economics 23. Urban Economics 24. Regional Economics 25. Poverty and Income Distribution

Sciences at Harvard University Richard A. and Susan F. Smith Campus Center 1350 Massachusetts Avenue, Suite 350 Cambridge, MA 02138 617-495-5315 gsas.harvard.edu Office of Diversity and Minority Affairs minrec@fas.harvard.edu gsas.harvard.edu/diversity Office of Admissions and Financial Aid admiss@fas.harvard.edu gsas.harvard.edu/apply

danbjork@fas.harvard.edu HARVARD UNIVERSITY Placement Director: Gita Gopinath GOPINATH@HARVARD.EDU 617-495-8161 Placement Director: Nathan Nunn NNUNN@FAS.HARVARD.EDU 617-496-4958 Graduate Administrator: Brenda Piquet BPIQUET@FAS.HARVARD.EDU 617-495-8927 Office Contact Information Department of Economics

HARVARD UNIVERSITY Placement Director: John Campbell JOHN_CAMPBELL@HARVARD.EDU 617-496-6448 Placement Director: Nathan Hendren NHENDREN@FAS.HARVARD. . Department of Economics, Harvard University (Jan. 2015-Jul. 2016) Economist, Monetary & Capital Markets Dept., International Monetary Fund (Dec. 2013-Dec. 2014) .

ECONOMICS 40 Chapter 1: The Principles and Practica of Economics 40 1.1 The Scope of Economics 41 Economic Agents and Economic Resources 41 Definition of Economics 42 Positive Economics and Normative Economics 43 Microeconomics and Macroeconomics 44 1.2 Three Principles of Economics